Built an AI Agent using Strands Agents SDK
Introduction
Built an AI Agent using Strands Agents SDK from Amazon Web Services (AWS) which calls my Kubernetes MCP server. Read more to find out…
Agents and MCP
AI Agents and Model Context Protocol are the most popular concepts in Gen AI now. Now I have created an AI Agent which calls MCP server to debug issues in my K8s cluster.
Recently I created a Model Context Protocol (MCP) server for Kubernetes read-only operations.
Strands Agents SDK
AWS has created an SDK for building AI Agents called Strands Agents SDK. Now it added support for MCP as well.
I used Strands Agents SDK and built an AI Agent which calls my K8s MCP server and debugs issues in my running K8s cluster.
This demonstrates the AI Agents ability to achieve goals by perceiving the environment, reasoning and acting upon it using available tools like MCP servers.
First impressions
I have built agents using Langchain/Langgraph before. Compared to that creating agents using Strands Agents is simple and straight-forward. Add to that you can now call MCP servers also.
Icing on the cake is that Strands Agents supports any LLM and not only restricted to Amazon Bedrock. Thanks to LiteLLM you can call most of the LLMs using this library. Strands also supports Ollama for calling LLMs running locally.
Really impressed with it.
Demo
Below is my detailed demo of my AI Agent:
What are you building with Strands Agents SDK?
Follow Me
If you are new to my posts, I regularly post about GenAI, AI Agents, MCP, AWS, EKS, Kubernetes and Cloud computing related topics. Do follow me on LinkedIn and visit my website (https://vijay.eu/posts) where I have all my previous posts at one place.
